Thursday, August 25, 2011

Game Theory Application in Pitching Sequences


game theory

noun
a mathematical theory that deals with strategies for maximizinggains and minimizing losses within prescribed constraints, as therules of a card game: widely applied in the solution of variousdecision-making problems, as those of military strategy andbusiness policy.


Through my time as a professional poker player in college, I dabbled and experimented and educated myself in any way possible to gain an edge on the competition. For those of you whom are not aware, poker has adapted (especially online poker) from old school degenerate gamblers who are dumb as rocks, to extremely intelligent and talented mathematicians who are ready and willing to exploit any edge they can identify. It has adapted beyond a game of cards, into a smaller, less complex version of the stock market.

One of the topics I found extremely interesting was game theory. In essence, what game theory teaches you, or attempts to teach you is the way to "solve" a particular set of problems. In reality, that's what any game is: a set of problems that must be solved as efficiently as possible to achieve the greatest result. As in poker, there is a few players who are so talented mathematically that they have "solved" heads-up limit hold'em to create a situation where no opponent can achieve an edge more than 50%. Incredible, but even a computer cannot beat these brilliant minds when playing optimally over a big enough sample size.

Baseball analysis is advancing at light speed. The terms the average fan uses now were advanced concepts 10 years ago. The thing that GM's are using now are ways to come as close to an optimal decision as possible, all things considered of course. They may not be right every single time, but they're right more often.. and that's all you can really do in business, in the markets, and in life. Using mathematics with an ever-increasing sample size is allowing the brilliant minds to undercover things the naked eye can't and won't see. 

About a month ago I was on a plane and thought of how many similar parts there are to poker, investing and baseball. So I proposed a question that I still don't quite have the answer to: Is it possible to apply game theory to pitch selection. And if so, are pitchers and/or pitching coaches currently using these methods in pre-game preparation and analysis of opposing pitchers? If a pitcher throws a 3-2 fastball too often, he's not optimally presenting enough of a portfolio of pitches and will likely be hit hard. Almost anybody can figure this out, but is there room for improvement and who is capable of doing this?

The first guy that comes to mind is Cliff Lee. He is mechanically sound, and doesn't give away what he is throwing. He is able to throw 5-6 different pitches and command/control them all with great proficiency. He often throws a pitch that you don't expect him to because he has the ability to locate every pitch in his arsenal. But is that feel, or does he apply math to his pitching sequences? I have no clue because I am not his pitching coach, but it's interesting to think about and if this is something pitching coaches and organizations have philosophically taken on.

What do you think?

No comments:

Post a Comment